Carbonyl sulfide (COS) emissions in two agroecosystems in central France

Carbonyl sulfide (COS) fluxes simulated by vegetation and soil component models, both implemented in the ORCHIDEE land surface model, were evaluated against field observations at two agroecosystems in central France. The dynamics of a biogenic process not yet accounted for by this model, i.e., COS emissions from croplands, was examined in the context of three independent and complementary approaches. First, during the growing seasons of 2019 and 2020, monthly variations in the nighttime ratio of vertical mole fraction gradients of COS and carbon dioxide measured between 5 and 180 m height (GradCOS/GradCO2), a proxy of the ratio of their respective nocturnal net fluxes, were monitored at a rural tall tower site near Orléans (i.e., a "profile vs. model" approach). Second, field observations of COS nocturnal fluxes, obtained by the Radon Tracer Method (RTM) at a sub-urban site near Paris, were used for that same purpose (i.e., a "RTM vs. model" approach of unaccounted biogenic emissions). This site has observations going back to 2014. Third, during the growing seasons of 2019, 2020 and 2021, horizontal mole fraction gradients of COS were calculated from downwind-upwind surveys of wheat and rapeseed crops as a proxy of their respective exchange rates at the plot scale (i.e., a "crop based" comparative approach). The "profile vs. model" approach suggests that the nocturnal net COS uptake gradually weakens during the peak growing season and recovers from August on. The "RTM vs. model" approach suggests that there exists a biogenic source of COS, the intensity of which culminates in late June early July. Our "crop based" comparative approach demonstrates that rapeseed crops shift from COS uptake to emission in early summer during the late stages of growth (ripening and senescence) while wheat crops uptake capacities lower markedly. Hence, rapeseed appears to be a much larger source of COS than wheat at the plot scale. Nevertheless, compared to current estimates of the largest COS sources (i.e., marine and anthropogenic emissions), agricultural emissions during the late stages of growth are of secondary importance.